6m in 2nd Nyquist zone

[Clifford Heath]

Has anyone tried receiving at 50MHz using an HL2? With a good BPF (and maybe gateware changes?) it should be possible to receive in the 2nd Nyquist zone.

Clifford Heath 

[Luca De Biase]

Hi Clifford,

can you explain a litlle more what do you mean with:  it should be possible to receive in the 2nd Nyquist zone?

As i know HL2 max frequency is 38.4 Mhz.

Thank you.

Luca

[Clifford Heath]

On 16 Nov 2024, at 12:45 pm, Luca De Biase <lucala...@gmail.com> wrote:

can you explain a litlle more what do you mean with:  it should be possible to receive in the 2nd Nyquist zone?

As i know HL2 max frequency is 38.4 Mhz.

Il giorno martedì 12 novembre 2024 alle 22:06:38 UTC Clifford Heath ha scritto:

Has anyone tried receiving at 50MHz using an HL2? With a good BPF (and maybe gateware changes?) it should be possible to receive in the 2nd Nyquist zone.

It’s just sampling theory. The HL2 samples at 76.8MHz, so when it looks at a signal at 38.4MHz, it sees exactly every half-cycle (this is a problem if the sample is always taken at the zero-crossing point, because it sees only DC!).

If the input signal is a 38.0MHz sine wave, each half-cycle sample falls a bit earlier. So if the positive half-cycle is exactly at the peak, the negative will be slightly before the peak, and this will rotate the sampling phase at 0.4MHz.

If the input signal is a 38.8MHz sine wave, each half-cycle falls a bit later. But the samples will appear to be exactly the same as for the 38.0MHz sine. All the frequencies from 38.4 up to 76.8MHz appear as images from 38.4 down to 0Hz. Increasing signal frequency results in a reduced frequency in the digitised signal. This is called an image frequency, in the second Nyquist zone.

Note that it’s very difficult to construct a filter that passes 38.0MHz but attenuates 38.8MHz; so there will always be overlap between the Nyquist zones. We normally wouldn’t quote a 76.8MHz ADC as capable of more that 40% of that, or 30.7MHz, and that’s true with the HL2 also .

Above 76.8 MHz, say at 77.2MHz, the successive samples are taken slightly further forward than one full cycle. The digitised signal appears to be at 400kHz. This is the 3rd Nyquist zone, which continues a further 38.4MHz up to 115.2MHz, at which point the received signal is indistinguishable from a 38.4MHz one.

Above that is the 4th Nyquist zone, where mirroring (image frequency) continues up to 153.6MHz, and so on.

So if the HL2 receiver is used (say) with a 40-70MHz band-pass filter, it will receive them as if they were at 36.8->6.8MHz. Happily, the 6m band is right in the middle, and should be easy to receive using the HL2. Just pay attention to the frequency reversal (and demodulate the opposite sideband!). 50MHz will be received at 26.8MHz, for example.

Transmitting might also be possible, because the reproduced waveform from the DAC also contains these higher frequencies.

Clifford Heath.

[Steve Haynal]

Hi Clifford,

Yes, this has been done. The AD9866 ADC is not designed for undersampling like some other ADCs, but it can do a passable job for 6M. I didn't want to include support for 6M because it complicates and adds expense to the frontend to switch RX filters and could introduce more noise. Instead, RF4 and RF5 are included on the HL2 for experimentation. One can remove the LPF composed of C53/C55/C56/C57/L10/L11 and insert a new filter between RF4 and RF5. This could be a switchable filter between the 1st and 2nd Nyquist zones. No one has made use of this experimental feature as far as I know. This is for RX only. Transmit may be a little more difficult as the AD9866 DAC has some filtering in the mode we are using.

Search this list for 6M. In particular, look for JI1UDD who did the initial 6M experiments with the original Hermes-Lite. SparkSDR still has support for HL2 and 6M.

https://github.com/ji1udd/Hermes-Lite/tree/6M

https://groups.google.com/g/hermes-lite/search?q=6M%20JI1UDD

73,

Steve

kf7o

[Luca De Biase]

Interesting, thank to both for the informations..